This forum is shutting down! Please post new discussions at community.naturalcapitalproject.org

Habitat Quality Model - linear vs exponential decay

heatherheather Member
edited February 2014 in Terrestrial Models
Hello
I am running the habitat quality model and I read in the documentation that you can choose if you want linear or exponential decay. I don't see anywhere in the model set up that allows this choice. The distance over which I am seeing degradation is very small compared to my max distances so I assume it defaults to exponential, but is there a place to change this?
Thanks!

Comments

  • DougDoug Administrator, NatCap Staff
    Hi @heather,

    We made the decision when moving from the ArcGIS version to the standalone version to drop linear decay and just use exponential. Currently it is not possible to switch out of exponential. Seeing how you would like to see this feature back, I will get a hold of our scientist behind this model and talk it over!

    Here's the link to the latest User's Guide and documentation online:

    User's Guide: Habitat Quality: Biodiversity

    There's a blurb under, Habitat Quality Standalone Beta, mentioning the drop in linear decay.

    Let me know if I can help in any other way!

    Doug
  • Hi Doug

    Thanks for your quick response. I didn't notice that when reading the documentation. What version would I need to go to in order to try the linear buffers? Is it still available for download?

    Also, I realize that as it is exponential the magnitude of degradation should drop rapidly but it seems to me to be too quick. For example, with a max distance of 150m, the degradation has dropped to zero by 30 m. Does this sound correct?

    Thanks,
    Heather
  • Hi Doug

    One more question - do you know why the option was removed?

    Thanks,
    Heather
  • DougDoug Administrator, NatCap Staff
    Hey Heather,

    Unfortunately the new standalone version was never developed with the linear decay option.

    Your example does sound a little strange. I'm going to ping a few of our developers and get their input. I'll also look into the reasoning for dropping linear decay!

    Thanks for your patience.
  • DougDoug Administrator, NatCap Staff
    Hi Heather,

    I just received some good feedback from the developers that hopefully will help. First, it was decided to drop linear decay because it seemed unnecessary and unrealistic. The idea being that usually nothing in nature has a linear decay.

    And to follow up with your decaying too early issue, did you properly set the half saturation constant? The half saturation constant should be set to half the highest degradation value of the first run. During the first run you can leave the half saturation at the default. This can be seen in the User Guide under the 8th data input.

    What is the resolution of your inputs, specifically the LULC?

    If this doesn't give you a proper degradation value please let me know. I'd be happy to check out your data to make sure their isn't a bug.

    Thanks!
  • Hi Doug

    Thanks for your response. I did adjust the half saturation constant after the first run. My highest degradation value was .16 so I set it to 0.08 and re-ran but the results were the same. My inputs are 1.5m resolution. We have very narrow linear disturbances so we need to use quite a high resolution.

    We have max distances of 15 m (degration occurs to around 5-8 m depending on the land cover) and 150 m (degradation occurs up to 30 m).

    These are the values I used for the 10 threats we included in our landscape. Do you see anything there that might be a problem? Perhaps the weights are having an effect?

    MAX_DIST WEIGHT
    0.015 0.25
    0.015 0.5
    0.15 0.5
    0.15 0.5
    0.15 1
    0.15 1
    0.15 0.25
    0.15 0.1
    0.15 0.5
    0.15 1

    Thanks again,
    Heather
  • DougDoug Administrator, NatCap Staff
    Hi Heather,

    I do believe the weights will indeed have an effect as well as if you provided any protected areas in a shapefile. Do you see the same degradation with weights of 1.0 as with weights of .25?

    I have notified some colleagues of this post and hope to hear their feedback soon!

    Thanks for your patience.
  • Hi Doug

    I finally had a chance to run the model using weights of 1.0 on all threats (same max distance). The results are the same - degradation is zero after about 30 m for the 150 m max distance. So it appears increasing the weights did not make a difference.
    I can see that a linear decay may not be particularly useful but it does seem that the exponential is decreasing too rapidly.

    Is it possible to find out if the decay rate I am seeing is expected based on the intention of the model? If so, it seems that to actually get a degradation effect up to 150 m I would need to play around with the max distance, increasing it until I see an effect, which is of course not ideal.

    Thanks,
    Heather


  • DougDoug Administrator, NatCap Staff
    Hi Heather!

    We just wrapped up our Annual Meeting last week and I had a user in one of our sessions bring up a similar problem that you are experiencing. We were able to walk through their very simple data and see if the equations were producing what we would expect.

    As it turns out, there may be a bug in the model where we decay the threats across space. I'll be actively investigating this issue in the coming week and will keep you updated on this thread. Thanks for your patience and help in debugging our tools and making them better.

    Doug
Sign In or Register to comment.